G Model PPEES-25156; No. of Pages 7 ARTICLE IN PRESS Perspectives in Plant Ecology, Evolution and Systematics xxx (2012) xxx–xxx Contents lists available at SciVerse ScienceDirect Perspectives in Plant Ecology, Evolution and Systematics j ournal homepage: www.elsevier.de/ppees Research article Delving into the loss of heterostyly in Rubiaceae: Is there a similar trend in tropical and non-tropical climate zones? a,b,∗ b b b V. Ferrero , D. Rojas , A. Vale , L. Navarro a Centre for Functional Ecology, Department of Life Sciences, Faculty of Sciences and Technology, University of Coimbra, PO Box 3046, 3001-401 Coimbra, Portugal b Department of Plant Biology and Soil Sciences, Faculty of Biology, University of Vigo, As Lagoas-Marcosende 36200 Vigo, Spain a r t i c l e i n f o a b s t r a c t Article history: Heterostyly is a specialised floral polymorphism consisting in the presence within the populations of two Received 18 May 2011 or three morphs that differ reciprocally in sexual organ position. The function of heterostyly has usually Received in revised form been related to the promotion of cross-pollination fostered by the perfect adjustment between pollina- 22 November 2011 tors and flower morphologies. Rubiaceae is the largest family in which this polymorphism is present. Accepted 28 November 2011 Nevertheless, just a few studies on the evolution of heterostyly have been carried out in this family. To investigate the appearance and maintenance of heterostyly we select the subfamily Rubioideae as Keywords: study group. Rubioideae occur in both tropical and temperate regions and since the tropics are known to BayesTraits contain higher biodiversity and greater ecological specialisation than temperate areas, we characterise Floral polymorphism the taxa as tropical, non-tropical or mixed distributed (when they are present in tropical and non-tropical Pollinator efficiency Spermacoceae areas) and explored whether the heterostyly, as a specialised system, is more stable in tropical regions than in other climates of the world. Ancestral nodes in Rubioideae present heterostyly, which also is maintained along most evolutionary lineages of this group. Although we do not find a significant correlation between the presence of het- erostyly and the climate zones along the whole subfamily, our results show that two of the main clades in the Spermacoceae alliance where heterostyly is lost are distributed in non-tropical areas or, at least, they are not restricted to tropical distributions. These results partially support the hypothesis that plant lineages when exposed to different pollina- tion scenarios may evolve towards divergent pollination systems and different degrees of specialisation. However, a more detailed analysis at the species level is suggested for future studies. © 2011 Elsevier GmbH. All rights reserved. Introduction Thomson, 1994; Nishihiro et al., 2000; Lau and Bosque, 2003; Ferrero et al., 2011). Still, most of the groups in which the origin Heterostyly is a genetically controlled floral polymorphism and evolution of heterostyly have been studied occupy temperate primarily characterised by the presence of two or three intra- ecosystems (e.g., Primulaceae, Mast et al., 2006; Lithospermeae, population morphs whose flowers differ reciprocally in the position Ferrero et al., 2009; Narcissus, Graham and Barrett, 2004; Pérez- of the stigmas and anthers (Barrett, 2002). The function of het- Barrales et al., 2006). erostyly has been seen as a safeguard against self-fertilisation and Following the seminal work of Darwin (1877) and the hypoth- inbreeding depression on one hand, and as a mechanism of pro- esis of Lloyd and Webb (1992), heterostyly is considered as a motion of cross-pollination between morphs on the other (Barrett, specialised system because of the perfect adjustment between 1992). Some models explaining the evolution of heterostyly con- pollinators and flowers required for efficient pollen transfer and sider pollinators as the selective force driving the appearance thereby, for the origin and maintenance of stylar polymorphism of reciprocal herkogamy (Lloyd and Webb, 1992). This hypothe- in the plants. One of the most common hypotheses in this topic sis is supported by different pollination systems (e.g., Stone and is the increasing specialisation from temperate to tropical regions (Johnson and Steiner, 2000). Comparisons of plant-pollinator net- works from different latitudes support this idea, at least for ∗ flowering plants (Olesen and Jordano, 2002). However, contrary Corresponding author at: Centre for Functional Ecology, Department of Life Sci- results have been reported in plant pollinator systems (Ollerton ences, Faculty of Sciences and Technology, University of Coimbra, PO Box 3046, and Cranmer, 2002; Corlett, 2004), as well as in other mutualistic 3001-401 Coimbra, Portugal. Tel.: +34 986 812628; fax: +34 986 812556. E-mail address: [email protected] (V. Ferrero). systems, suggesting that this hypothesis may vary depending on 1433-8319/$ – see front matter © 2011 Elsevier GmbH. All rights reserved. doi:10.1016/j.ppees.2011.11.005 Please cite this article in press as: Ferrero, V., et al., Delving into the loss of heterostyly in Rubiaceae: Is there a similar trend in tropical and non-tropical climate zones? Perspect. Plant Ecol. Evol. Syst. (2012), doi:10.1016/j.ppees.2011.11.005 G Model PPEES-25156; No. of Pages 7 ARTICLE IN PRESS 2 V. Ferrero et al. / Perspectives in Plant Ecology, Evolution and Systematics xxx (2012) xxx–xxx the category of interactions and on the group of taxa under study Rydin et al., 2009). We randomly selected one species per genus (Ollerton and Cranmer, 2002). among the set of species for which most sequences were available In this study we propose, for the first time, heterostyly as a in GenBank. Taxa names were assigned following the Interna- scenario where to test the above-mentioned prediction. Tropical tional Plant Names Index (http://www.ipni.org/index.html). We forests are known to contain a higher diversity of species than used the taxonomic treatment of Psychotrieae and Spermacoceae any other terrestrial ecosystem (Whitmore, 1998), since resources alliances following Bremer and Manen (2000). In the case of para- split more finely among a greater number of species (McArthur, phyletic genera, we specified the group considered in the footnotes 1972; Janzen, 1973). The predictable climate in the tropics led to (Appendix S1). As the evolution of the heterostyly was assessed a greater importance of biotic interactions (Dobzhansky, 1950). at the genus level, we characterised a genus as heterostylous if at In these ecosystems we thus could expect to find more obligate least one of its species presented heterostyly. Such characterisation mutualisms than in temperate forests because tropical forests are was based on Bremekamp (1952), Ganders (1979), Keegan et al. stable environments where environmental perturbations to popu- (1979), Wyatt and Hellwig (1979), Steyermark (1988), Robbrecht lation sizes are hypothesised to be minimal (May, 1973; Farnworth (1988, 1993), Tange (1997), Rutishauser et al. (1998), Western and Golley, 1974). In temperate zones, oscillations in ecological Australian Herbarium (1998–), Delprete and Boom (1999), Taylor and environmental conditions are more frequent and drastic than and Clark (2001), Malcomber (2002), Naiki and Nagamasu (2003), in tropical areas (Fischer, 1960). For this reason, we can expect Kårehed and Bremer (2007), Puff (2007), eFloras (2008), Aluka that greater possible variations in pollinator arrays in temperate database (2000–2010), Consolaro (2008), García-Robledo and Mora environments could affect the maintenance of the heterostylous (2008), Sonké et al. (2008), Groeninckx et al. (2009), Neupane condition (e.g., Narcissus papyraceus, Pérez-Barrales et al., 2009) to et al. (2009) and Rydin et al. (2009). To characterise the genera a greater extent than in tropical systems. as tropical, non-tropical or mixed, we used the world checklist of When testing this kind of hypothesis, a proper phylogenetic selected plant families: Royal Botanic Gardens, Kew (available at: framework is required together with an adequate characterisation http://apps.kew.org/wcsp/home.do) and the updated World Map of the floral morphology. Rubiaceae are one of the largest fami- of the Köppen–Geiger climate classification (Kottek et al., 2006). lies of flowering plants with more than 13,100 species, 611 genera The climate classification combines temperature and precipitation and 3 subfamilies (Robbrecht and Manen, 2006; Govaerts et al., conditions. Main climates are (A) Equatorial (minimum tempera- ◦ 2006; Bremer and Eriksson, 2009). Rubioideae and, particularly, ture ≥+18 C), (B) Arid (accumulated annual precipitation <10 mm), − ◦ the Spermacoceae alliance (sensu Bremer and Manen, 2000), are (C) Warm temperate (minimum temperature above 3 C and ◦ ◦ distributed worldwide and possess high diversity of life forms and below +18 C), (D) Snow (minimum temperature <−3 C) and (E) ◦ flower morphology (Bremer and Manen, 2000). Heterostyly is well Polar (maximum temperature <+10 C). We defined the taxa as known to be present especially in this subfamily (e.g., Keegan et al., tropical when they were distributed in regions classified as (A); 1979; Sobrevila et al., 1983; Ree, 1997; Faivre and McDade, 2001). non-tropical when distributed in (B), (C) and/or (D); and, as mixed Rubioideae have been subjected to numerous phylogenetic stud- when distributed in (A) and other areas. We did not